Computer vision technology-based red tide real-time monitoring method

Abstract

The invention discloses a computer vision technology-based red tide real-time monitoring method. Through an interactive segmentation algorithm (a GrabCut algorithm) for extracting a foreground targetfrom a complex background, and by utilizing color space conversion, the influence of light ray conversion, water surface reflection and water surface floating objects in a monitored water area image can be resisted; a water surface foreground image can be efficiently and accurately obtained; the occurrence of a red tide can be accurately and quickly judged and an alarm is timely given; the cyclesof manual observation and sampling detection are greatly shortened; the limitations of a conventional detection method in personnel, time and space are broken through; related departments are facilitated to quickly adopt control measures; and large-area flooding of the red tide is avoided.

Description

technical field [0001] The invention belongs to a red tide real-time monitoring method, in particular to a red tide real-time monitoring method based on computer vision technology. Background technique [0002] Red tide is an abnormal phenomenon in the marine ecosystem. It is caused by the explosive proliferation of red tide algae in a specific environment. Red tide generally occurs in lakes and adjacent sea waters. The traditional method of manual on-duty sampling detection has problems such as long sampling period, low frequency, time-consuming and laborious, and cannot detect the occurrence of red tide in the first place, resulting in serious damage to marine ecology. Computer vision technology is a simulation technology of computer to biological vision. Various imaging systems are used to replace biological visual senses as a means of inputting information, and computers replace biological brains to complete information processing and interpretation. As an important bra...

AI Technical Summary

Problems solved by technology

However, since red tides often occur in areas with complex environments such as lakes and offshore areas, they are subject to relatively large environmental disturbances, such as light changes, water surface reflections, and the influence of floating objects in the water, etc.

Therefore, so far there is no relevant report on the real-time monitoring method of red tide based on computer vision technology

Images